Site: Jornada Basin LTER
Landscapes at the Jornada LTER site show complex patterns related to vectors that redistribute resources: water flows from small mountain ranges to the basin (lower left), wind moves soil and nutrients (upper right), cattle concentrate impacts near water sources (center), and human activities occur on-site via roads (center) and on adjacent private land (lower right).
USDA ARS, Jornada Experimental Range archive

Desertification is the shift from perennial grasslands to shrublands that occurs globally to impact nearly 40% of the Earth’s land surface and a fifth of the world’s human population. Desertification results from interactions between human activities (such as livestock overgrazing) and prolonged drought. However, these broad-scale drivers are insufficient to explain variability in dynamics at finer scales. Long-term research at the JRN is providing new insights in understanding spatial and temporal variability in the causes and consequences of desertification.

Large areas of the Jornada were formally perennial grasslands that, beginning in the late 19th century, were invaded by shrubs and have since succumbed to desertification processes. Capitalizing on this history of vegetation change, JRN researchers developed a resource-redistribution conceptual model in the early 1990s that incorporates variation in abiotic factors, disturbance regimes, and, most importantly, plant-soil feedbacks that work synergistically to enhance desertification processes. Initial research at the JRN focused on interactions among plants (grasses or shrubs) and adjacent non-vegetated areas. During desertification, loss of grass cover by overgrazing and drought leaves soil exposed to erosion by wind and water. These erosive forces redistribute soil nutrients from bare areas to beneath shrub canopies to result in “islands of fertility”. Through this process, the spatial distribution of nutrients changes from homogeneous in grasslands to fragmented and heterogeneous in shrublands. Thus, desertification entails both a biotic and abiotic restructuring of a system that is often very difficult to reverse. Consequences of desertification include loss of forage production, biodiversity, and soil nutrients locally, and a reduction in air quality, wind and water erosion of topsoil, and loss of water to overland flow at landscape to regional scales.

More recently, JRN research has expanded from a focus on the plant – interspace scale to explore the implications of resource redistribution across a range of interacting scales. JRN researchers developed a multi-scale model that emphasizes quantifying transfers of materials across scales for different vectors (wind, water, animals) within and among ecosystem types. These vectors act to connect spatial units across landscapes; this connectivity can either accelerate or attenuate rates of desertification at finer and broader scales. These vectors also vary in their importance at different scales in different locations. For example, the susceptibility of sandy soils to broad-scale wind erosion and deposition interacts with the redistribution of soil and nutrients by water at plant-interspace scales. Small and large animals move across locations as herbivores, but also influence system dynamics through seed dispersal, nutrient redistribution, and soil compaction. Understanding how these biotic effects interact with physical processes of wind and water erosion to generate observed dynamics at multiple scales is providing critical information to future predictions.

This body of research by JRN scientists on the causes and consequences of desertification is applicable globally, and has demonstrated that a spatial accounting of connections and interactions among landscape components is needed to extract trends and signals in ecosystem drivers and dynamics from what had previously been regarded as noise.

Graph for
Jornada landscapes were dominated by perennial grasses in 1858. By 1915, 63% was shrub-dominated, and almost all of the area (92%) was shrubland in 1998.
Modified from Gibbens et al. 2006.